The Anatomic Substrate of Postmyocardial Infarction Ventricular Tachycardia

After an MI, the tissue can be broadly divided into three zones: the dense scar, the surrounding live myocardial tissue, and the intervening “border zone.” It is important to note that this border zone is not necessarily physically located only at the periphery of the scar, but is rather located at any of the interfaces between the normal tissue and dense scar. In this border zone, electrically-active live myocardial fibrils are interspersed among the bed of infarcted, fibrotic tissue. These fibrils are characterized by abnormal electrophysiologic properties including slower conduction velocity and decreased cell-to-cell electrical coupling (e.g., because of altered Connexin 43 activity at the gap junction). As with reentrant circuits located in other regions of the heart, the initiation of VT is dependent on the development of unidirectional block and slow enough conduction to allow the recovery of excitability of the initially blocked region to initiate a self-perpetuating reentrant circuit. The initiators of scar-related VT are not well understood. Presumably, a well-timed premature beat or series of premature beats arise as a result of triggered activity from discrete regions of the heart, and this allows for the unidirectional block and slow conduction required to initiate reentrant VT.

Once initiated, to maintain the reentrant circuit, the wavelength of the tachycardia circuit must be short enough, or the path of myocardial circuit long enough such that the wave front is constantly encountering excitable tissue. This can occur because of either (1) an anatomically-determined circuit of the appropriate length or (2) a partial anatomic barrier combined with a functional barrier. For example, a functional barrier may result from ischemia, electrophysiologic changes resulting from treatment with antiarrhythmic drugs, or electrolyte and pH changes ( Fig. 32.2 ). The anatomic compartmentalization combined with altered cell-to-cell electrical coupling of the diseased tissue sets the stage for local micro- (or macro-) reentrant circuits that result in VT and have the potential to culminate in VF.

The importance of the excitable gap to maintain a reentrant tachycardia circuit. A, The wave front traverses within the scarred tissue along a surviving tract of myocardial tissue. Conduction through this pathway is slow because of a number of potential factors including the arrangement of the myocardial fibers (side-to-side instead of end-to-end), alterations in gap junctions between myocardial fibrils, a meandering path of the tract, and slow conduction velocity at certain regions (e.g., areas of extreme wave front curvature). The wavelength of the circuit is short enough that the leading edge of the wave front constantly encounters excitable myocardial tissue. This “excitable gap” allows the circuit to perpetuate and manifest as ventricular tachycardia (VT). B, The wavelength of the tachycardia circuit is longer than the tissue tract that it must follow. The leading edge of the wave front encountered refractory tissue so the circuit extinguished, and VT was not maintained. C, However, functional block can supervene in certain situations such as ischemia, increased heart rates, administration of drugs that alter conduction velocity or ventricular repolarization, electrolyte changes, acid-base imbalances, etc. In this situation, the combination of functional block to the preexisting anatomical block creates an excitable gap, allowing for sustained VT.

Surgical Experience With Postmyocardial Infarction Ventricular Tachycardia

The approach to ablation of unstable VTs developed directly from the extensive experience since the late 1970s with surgical modification of the arrhythmogenic substrate in post-MI patients. Because the reentrant circuit is most often located in the subendocardium at the junction of normal and scarred myocardium, the initial surgical experience with simple aneurysmectomy was disappointing. However, two effective general strategies were developed over time: (1) subendocardial resection—involving surgical removal of the subendocardial layer containing the arrhythmogenic substrate in this border zone; and (2) encircling endocardial ventriculotomy—consisting of the placement of a circumferential surgical lesion through the border zone, presumably interrupting potential VT circuits. Because of its distinct advantage in destroying myocardial cells without disrupting the fibrous stroma, cryoablation has also been used both as a stand-alone intervention during surgery, and as an adjunct to subendocardial resection. Encircling cryoablation is an efficacious procedure incorporating cryoablation into the concept of an encircling endocardial ventriculotomy. When performed at experienced centers, the long-term freedom from malignant VT/VF after surgery is more than 90% ( Fig. 32.3 ).

Surgical substrate modification to eliminate ventricular tachycardia (VT). The border between the normal and infarcted/aneurysmal wall contains a stylized VT circuit—predominantly endocardial, and partially intramural. The surgical procedures, subendocardial resection and ventriculotomy, are thought to either remove or transect critical endocardial portions of the VT circuit, respectively. During catheter ablation, the border zone is mapped using the electroanatomic mapping system, the putative exit site of the VT is identified by pace mapping, and catheter-based linear lesions are placed in an attempt to interrupt the circuit. Because mapping is performed during sinus rhythm instead of during VT, this allows for greater patient safety and comfort.

Modified from Miller J, Rothman SA, Addonizio VP. Surgical techniques for ventricular tachycardia ablation. In: Interventional Electrophysiology . Baltimore, Md: Williams & Wilkins; 1997:641–684.

It is interesting to note that in the initial surgical experience, intraoperative mapping was performed to help guide the surgical resection. After open surgical bypass, multielectrode plaques were used to precisely identify the origin of the VT. This area of endocardium was either surgically removed, or surgically transected using a scalpel blade. However, VT surgery then evolved such that in many cases, equivalent results were obtained by visualizing the scar and either simply resecting it, or by placing surgical cryoablation or laser ablation lesions along its border. These “empiric” lesions are thought to eliminate critical portions of the circuit and thus render VT noninducible ( Fig. 32.4 ).

Electroanatomic mapping of the effect of arrhythmia surgery. A, Left ventriculography reveals a large inferobasal aneurysm in the setting of three-vessel coronary artery disease and clinical ventricular tachycardia (VT). During a presurgical electrophysiology study, programmed ventricular stimulation revealed easily-inducible VT. B, Left ventricular (LV) electroanatomic mapping was performed during sinus rhythm. The bipolar voltage amplitude maps shown in right anterior oblique caudal (left) and left lateral-caudal (right) projections reveal a large inferobasal aneurysmal scar with electrograms containing abnormal fractionated and late potentials (not shown). C, During surgery, the LV was opened through the aneurysm, the aneurysm was resected, cryoablation was applied to the margins of the scar, and the ventricle was closed with the support of a patch. D, Months after the surgery, a repeat electrophysiology study revealed (1) a smaller homogeneous scar without evidence of fractionated and late potentials (right posterior oblique and inferior projections on left and right , respectively); (2) a more favorable ventricular geometry without an aneurysmal component; and (3) no inducible VT with programmed ventricular stimulation. The color range is set such that purple represents normal tissue (>1.5 mV), red the most severely disease tissue (<0.1 mV), and gray represents pure scar with no identifiable electrical activity. AV , Aortic valve; MV , mitral valve.

The effect of arrhythmia surgery on the myocardial substrate was examined in a study of 18 patients undergoing successful subendocardial resection procedures. These patients had all previously sustained anterior wall MIs and manifested multiple morphologies of drug-refractory monomorphic VT. During the operative procedure, a 20-electrode rectangular plaque array was used to obtain electrical data from the apical septum during VT as well as during normal sinus rhythm immediately before and immediately after resection of subendocardial tissue ( Fig. 32.5 ). Electrograms (EGMs) could be compared from 298 of 360 (83%) of the electrodes. Before resection, split EGMs were present in 130 (44%) and late potentials in 81 (27%) of the recordings. However, the postresection recordings revealed a complete absence of the split EGMs, as well as elimination of all of the previously recorded late potentials. The mean EGM duration decreased from 112 ± 38 to 65 ± 27 ms, primarily caused by the loss of these split and late potentials. Histologic studies revealed that the subendocardial tissue removed in this procedure contains bundles of surviving muscle fibrils separated by dense connective tissue. These data suggest that the direct effect of the subendocardial resection procedure is to eliminate the tissue containing these abnormal EGM components.

The electrophysiologic effect of subendocardial resection surgery. Recordings were made using a 20-bipole plaque array (A) before and (B) after resection as well as after replacement of (and recording through) the resected tissue specimen (C). The dotted line in (A) denotes the end of the QRS complex. The split and late electrogram (EGM) components ( arrows ) before resection are absent in the postresection recordings. In addition, most channels show an increase in amplitude of the remaining early EGM component, which maintains the same general morphology as before resection. After replacement of the specimen, these early EGMs appear similar to those obtained before resection, but note the absence of split and late electrograms (channels 5 and 20 did not record properly).

From Miller JM, Tyson GS, Hargrove WC, 3rd, Vassallo JA, Rosenthal ME, Josephson ME. Effect of subendocardial resection on sinus rhythm endocardial electrogram abnormalities. Circulation . 1995;91(9):2385-2891.

The surgical experience provided several important lessons that are relevant for modern catheter ablation of VT: (1) critical portions of the VT circuit reside on the endocardial surface of the scar (allowing access via a percutaneous endoluminal approach); (2) the majority of VTs exit from the border of the scarred myocardium; (3) during normal sinus rhythm, the “anatomy” of the scar can be delineated by certain criteria distinguishing abnormal endocardial EGM—low voltage amplitude, prolonged EGM duration, and the presence of late potentials; and (4) empiric disruption of the arrhythmogenic substrate containing these abnormal fractionated, discrete, split or late potentials in this “border zone” area can eliminate VT.

Other Scar-Related Ventricular Tachycardias

Reentrant VT also occurs from myocardial scar in the setting of other forms of cardiac pathology such as dilated cardiomyopathy (DCM). Histologic studies of myocardial tissue from patients with DCM have revealed multiple patchy areas of interstitial and replacement fibrosis and myofibrillar disarray with variable degrees of myocyte hypertrophy and atrophy. A necropsy study in patients with idiopathic DCM revealed that despite a relative paucity of visible scar (14%), a high incidence of mural endocardial plaque (69%–85%) and myocardial fibrosis (57%) was found. As with post-MI VT, the mechanism of VT related to DCM is also most commonly reentrant and is related to scarred substrate. Unlike post-MI scar, there is no predilection for an endocardial location. Delayed contrast enhanced magnetic resonance imaging (MRI) has demonstrated that scar in DCM may also be found in the epicardium and midmyocardium. Therefore a combined epicardial and endocardial ablation is often necessary to successfully abolish VT. In addition, in our experience, and that of other investigators, by electroanatomic mapping, the scar tends to be predominantly localized to the basal regions of the left ventricle (LV) and ventricular septum. Clinical studies of VT mapping and ablation in the setting of DCM have revealed that a substrate-based approach is also useful to eliminate these arrhythmias.

Defining the Scar With Electroanatomic Mapping

Definition of the ventricular scar and the arrhythmogenic areas in and around the scar with electroanatomic mapping is critical to understanding potential sites for ablation ( Table 32.1 ). Based on surgical mapping studies in patients with post-MI VT, there are several EGM characteristics during sinus rhythm that help to distinguish abnormal myocardial tissue, including low voltage amplitude, prolonged EGM duration, and fractionation with late and split potentials ( Fig. 32.6 ). Most electroanatomic maps today delineate abnormal ventricular tissue using a peak-to-peak bipolar EGM amplitude cutoff of 1.5 mV ( Fig. 32.7 ). This is largely based on the study by Marchlinski et al. that measured left and right ventricular bipolar voltage in six patients without structural heart disease (average age 37 years, five men). The distribution of bipolar EGM signals measured with a 4-mm-tip catheter showed that 95% were greater than 1.55 mV. A subsequent study in seven patients with structurally normal hearts that measured LV EGM amplitude with a 4-mm-tip catheter corroborated these findings.

Abnormal electrograms (EGMs). A, An example of a normal EGM is recorded from a catheter placed at the right ventricular apex (RVA); it is characterized by high amplitude, short EGM duration, and no electrical activity noted after the end of the QRS complex ( red line ). On the mapping catheter, the EGM is low amplitude, fractionated, and also has a late component that continues past the end of the QRS complex. B, A late potential is shown on the mapping catheter. There is a sharp component that is after the end of the QRS complex ( red line ) . C, When pacing from a location at which a late potential was recorded, there is a delay between the timing of the stimulus to the beginning of the QRS complex ( red line ) . This latency represents the time required for the wave front to traverse from the surviving myocardial tissue within the infarct to the normal myocardium.

Substrate mapping of the left ventricular (LV) endocardium. Electroanatomic mapping of the LV in this patient revealed a large anterior and apical myocardial infarction based on bipolar voltage amplitude. A, Right anterior oblique projection; B, left anterior oblique; C, left lateral; and D, inferior views. The color range in the bipolar voltage are 0.5–1.5 mV; purple and red represent normal and severely diseased tissue, respectively.

Bipolar voltage amplitude values lower than 0.5 mV are typically arbitrarily defined as dense scar, and values between 0.5 to 1.5 mV as abnormal tissue, which represents the border zone ( Table 32.2 ). This tissue often contains surviving fibrils that can be identified as split/late potentials in sinus rhythm or mid-diastolic potentials during VT. Imaging studies using contrast enhanced computed tomography (CT) combined with positron emission tomography (PET) and MRI have evaluated and largely validated this classification based on bipolar voltage amplitude criteria. However, it is important to keep in mind that the commonly used threshold for abnormal bipolar EGM amplitude is empirically derived and not necessarily reflective of histologic scar in a particular patient. Numerous factors may affect EGM features during an individual case, including angle of the electrical wave front compared with the catheter, electrode size, interelectrode spacing, and signal filtering. Patient-specific factors such as ventricular hypertrophy may call for upward adjustment of the cutoff values used for abnormal EGMs. Chronic pacing could also affect bipolar EGM amplitude, and a study in 11 post-MI patients revealed that 8% of sites had a bipolar EGM amplitude that was “reclassified” from abnormal (≤1.5 mV) to normal (>1.5 mV) or vice versa when alternating between ventricular and atrial pacing. Although substrate maps generated using standard definitions of bipolar EGM amplitude generally provide robust representations of the infarct morphology, it is important to allow for flexibility in scar definitions based on the characteristics of a particular case.

In some patients, scar may be mainly midmyocardial and the bipolar voltage maps from the endocardium and epicardium may underestimate the amount of scar, or on occasion may display completely normal voltage. Endocardial unipolar EGM amplitude is helpful to identify the presence of intramural or epicardial substrate. The presence of intramural or epicardial scar is suggested by the presence of an endocardial unipolar voltage lower than 8.3 mV in the left ventricle and lower than 5.5 mv in the right ventricle ( Fig. 32.8 ). Endocardial unipolar voltage maps may be used as a surrogate for bipolar voltage maps to identify scar and together with pace mapping can be used to identify VT exit sites to guide ablation ( Fig. 32.9 ).

Abnormal unipolar voltage map in a patient with nonischemic cardiomyopathy. A, In this patient with nonischemic cardiomyopathy and recurrent ventricular tachycardia (VT), the endocardial bipolar voltage map of the left ventricle (LV) (left anterior oblique view) displayed normal voltage. However, there was a location (black point) with normal bipolar voltage and a late potential ( arrow ) . B, The unipolar voltage map (left anterior oblique view) shows an abnormal voltage lower than 8.3 mV over the lateral wall of the LV suggesting the presence of intramyocardial and epicardial scar. C, Indeed the epicardial bipolar voltage shows a large anterior and lateral wall scar. Substrate-based ablation was performed epicardially. As the phrenic nerve (blue points) coursed through the area of scar, a balloon catheter was placed epicardially to separate the ablation catheter from the phrenic nerve during ablation. The color range in the bipolar voltage is 0.5–1.5 mV; purple and red represent normal and severely diseased tissue, respectively. The color range in the unipolar voltage is 0–8.3 mV; purple represents normal tissue. Black points , late potentials; pink points , fractionated potentials; green points , pace mapping sites; red points , ablation sites; blue points , sites of phrenic nerve capture with pacing.

The utility of unipolar voltage maps in guiding radiofrequency ablation. A, Pacemapping identified an area ( blue point , white circle ) that had a good pace match to the clinical ventricular tachycardia (VT). The endocardial and epicardial bipolar voltage in this area was normal (left lateral views). However, radiofrequency ablation ( red points ) was performed endocardially because of the good pace match and abnormal unipolar voltage in this region B, The endocardial unipolar map (left lateral view) shows that this area was abnormal because of the presence of a voltage lower than 8.3 mV. Presumably, the myocardial scar was intramyocardial rather than endocardial or epicardial. The color range in the bipolar voltage are 0.5–1.5 mV; purple and red represent normal and severely diseased tissue, respectively. The color range in the unipolar voltage are 0–8.3 mV; purple represents normal tissue. Pink points , fractionated potentials; green points , pace mapping sites; red points , ablation sites

Mapping based on bipolar EGM duration (using the electroanatomic mapping system’s double annotation function) has also proven to be capable of delineating the scarred myocardium (>50 ms or >100 ms in porcine and human ventricles, respectively). Because one must manually annotate the points to generate an EGM duration map, the practical utility of this approach may be limited. However, it is of interest that certain points that appear to be of artifactually-low voltage amplitude because, for example, of poor catheter–tissue contact, are often shown to be of normal EGM duration. We do not routinely perform EGM duration maps; however, the information is often used to help determine the validity of low amplitude signals during 3-dimensional bipolar voltage mapping. It is also important to note that any single mapped site with voltage amplitude greater than 1.5 mV (or EGM duration <50 ms or <100 ms in porcine and human ventricles, respectively) is not necessarily normal. However, although such a site may be incorrectly identified as normal (or abnormal) based on EGM criteria, constructing a map using a large number of locations has the effect of minimizing the effects of these “outliers” and identifying the scar and border zone in a clinically useful manner.

Pace Mapping for Ventricular Tachycardia Exit Sites

By definition, a reentrant rhythm is always depolarizing some quantity of myocardial tissue. Because the small mass of myocardial tissue in the protected myocardial channels within the scar contribute negligibly to the surface QRS, the QRS complex of a VT initiates when the wave front of activation emanates from the border of the scar. Accordingly, once the myocardial scar is defined, a brief examination of the surface QRS morphology of the target VT(s) can generally regionalize the VT exit site to a scar border. The vast majority of VTs in the setting of structural heart disease originate from the left ventricle. Accordingly, a left bundle branch block-like morphology in lead V ₁ indicates that the VT is exiting from the LV septum, or very rarely, from the right ventricle proper. The remaining VTs exiting from other regions of the left ventricle typically have a right bundle branch block morphology in lead V ₁ . However, it should be noted that a right bundle branch morphology VT can still have a septal exit site—a situation in which the frontal plane axis is typically leftward (positive in leads I and aVL).

The electrocardiogram (ECG) frontal plane axis can help differentiate an anterior versus inferior exit; the former is characterized by an inferiorly directed QRS axis with positive complexes in leads II, III, and aVF, and the latter is characterized by a superiorly directed QRS axis (negative II, III, and aVF). An apical exit is characterized by predominantly negative QRS complexes in the precordial leads, while basal exit sites tend to be predominantly positive in these leads. Although these rules are helpful, a number of factors can influence the QRS complex in any given patient including the size and location of the myocardial scar, the orientation of the heart in the thorax (horizontal vs. vertical), and intrinsic conduction system disease that can modify the wave front of activation.

Based on the ECG morphology of the VT, pace mapping is performed at the suspected border(s) of the scar during normal sinus rhythm to precisely localize the exit point. Unlike during VT, pacing during sinus rhythm results in omnidirectional spread of activation—which one may expect to result in a different paced-QRS morphology than the VT-QRS morphology even when pacing from the proper exit site. However, the optimal paced-QRS morphology is often only slightly different than the target VT-QRS morphology ( Fig. 32.10 ). This is likely because when pacing at a scar border, activation proceeding into the scar is slower and contributes little to overall ventricular activation when compared with the “orthodromic” wave front that rapidly emanates in the opposite direction into normal tissue. Not surprisingly, less optimal matches of the pace map exit sites are found when pacing along the borders of smaller scars as opposed to larger scars.

Pace mapping along scar border to identify ventricular tachycardia (VT) exit sites. Electroanatomic mapping of the left ventricular endocardium in a patient with prior anterior myocardial infarction was performed. The bipolar voltage maps demonstrated a large anterior, septal, and apical scar ( middle upper panel —superior left anterior oblique projection, middle lower panel —left lateral projection). Pace mapping was performed along the scar border to identify exit sites for two VTs (VT1, VT2). Pacing at the superior anteroseptal portion of the scar identified an excellent pace match to VT1 ( middle upper panel — red circle, green point ). Pacing from the lateral basal aspect of the scar yielded a good pace match to VT2 ( middle lower panel — red circle, green point ). These sites were subsequently ablated as they represented exit sites for the VTs. The color range in the bipolar voltage are 0.5–1.5 mV; purple and red represent normal and severely diseased tissue, respectively. Black points , late potentials; pink points , fractionated potentials; red points , ablation sites.

Pace mapping is also affected by the rate, stimulus strength, and electrode polarity during pacing. At faster pacing rates, the “antidromic” wave front of activation into the scar may contribute less to the QRS morphology than during slow pacing. In addition, ventricular repolarization may fuse into and modify the QRS morphology during faster pacing rates. Because it is difficult to predict the effect of these variables, pacing is ideally performed at a rate similar to the target VT rate. By presumably capturing more distant (that is, far-field) tissue, increasing the stimulus strength can also affect the QRS morphology. We typically start pacing at low output and increase the output until several QRS complexes are captured in succession. It is interesting to note that if multiple QRS morphologies are seen at varying outputs, this is indicative of a protected region (or channel) of tissue. That is, at the lower output, only this region is captured by pacing, whereas at a higher output, the far-field tissue is also captured. In the ideal situation, pacing would be performed using unipolar pacing so that only the distal electrode can stimulate the myocardium and one can avoid inadvertent pacing by the proximal electrode. But unipolar pacing typically results in a larger stimulus artifact that can preclude accurate QRS morphology interpretation. And from a practical perspective, it is unusual for bipolar and unipolar pacing to be of markedly different morphologies—likely because the ablation catheter is typically not parallel to the tissue surface resulting in the proximal electrode not being in contact with the tissue.

Imaging to Characterize the Ventricular Tachycardia Substrate

When using cardiac imaging to identify locations with structural characteristics that suggest sites of origin for VT, the distribution and magnitude of myocardial scar is typically the most useful information to be gained. Cardiac magnetic resonance is well-suited to identify scar based on its spatial, temporal, and contrast resolution. Both severe wall thinning and greater magnitude of scar according to delayed contrast enhancement correspond with lower voltage areas defined by electroanatomic mapping, and critical isthmus sites for VT ablation have been shown to correlate with MRI-defined scar in ischemic cardiomyopathy and nonischemic cardiomyopathy. MRI may be especially helpful to identify VT substrate that is mid-myocardial or epicardial. In a study of 77 patients with scar-related VT, MRI identified septal or epicardial scar in 43% of patients with nonischemic cardiomyopathy and in 6% of ischemic cardiomyopathy patients. Among the 11 patients with subepicardial scar according to the pattern of hyperenhancement, all underwent successful ablation at an epicardial site. A major advantage of MRI is the lack of ionizing radiation. However, its use remains limited in many patients with implantable defibrillators, who comprise a large proportion of patients for VT ablation. Specifying the amount of signal intensity by contrast enhancement and the amount of wall thinning that corresponds to scar depends on parameters set during processing of the study, so that assessments of scar magnitude are generally not comparable between patients.

Multidetector computed tomography (MDCT) has outstanding spatial resolution, approximately 0.5 to 0.625 mm, and can be used to delineate important landmarks such as the coronary arteries and the phrenic nerve before ablation procedures. This modality has also been applied to identification of substrate for VT ablation, particularly among patients with implanted devices that preclude MRI. Areas with wall thickness less than 5 mm detected by contrast-enhanced MDCT are correlated with voltage lower than 1.5 mV by electroanatomic mapping, especially on the endocardial surface. More severe wall thinning less than 2 mm is sensitive and specific for transmurality of VT substrate, such that late and fragmented potentials appear on the endocardium in patients with postmyocarditis cardiomyopathy and on the epicardium in patients with ischemic cardiomyopathy. Hybrid imaging with PET/CT has been used to augment scar localization, and delayed-enhancement imaging may significantly increase the sensitivity of MDCT for detection of scar according to electroanatomic mapping.

Intracardiac echocardiography (ICE) has a unique role during VT ablation because it is used real-time, typically through an 8 or 10 F ultrasound catheter positioned in the right atrium or right ventricle. ICE has been used to identify the location and degree of transmurality associated with myocardial scar. One study of 17 patients with scar-related VT showed a high correlation between scar area according to wall motion abnormalities by ICE and areas identified with electroanatomic mapping. Increased echo intensity by ICE imaging has also been shown to identify scar, even differentiating epicardial scar on the LV lateral wall.